Add Jukebox IE

[youtube-dl] / youtube_dl / extractor / youtube.py

diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
index a7cb88b5a3e46e42f35f357573319f65af07da5f..de653cb3d24e6318f157d5177b87c6b09b913bef 100644 (file)
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@@ -1,12 +1,11 @@
 # coding: utf-8
-from __future__ import absolute_import
 
 import json
 import netrc
 import re
 import socket
 
-from .common import InfoExtractor
+from .common import InfoExtractor, SearchInfoExtractor
 from ..utils import (
     compat_http_client,
     compat_parse_qs,
@@ -59,7 +58,7 @@ class YoutubeIE(InfoExtractor):
         '18': 'mp4',
         '22': 'mp4',
         '37': 'mp4',
-        '38': 'video', # You actually don't know if this will be MOV, AVI or whatever
+        '38': 'mp4',
         '43': 'webm',
         '44': 'webm',
         '45': 'webm',
@@ -130,12 +129,13 @@ class YoutubeIE(InfoExtractor):
         """Indicate the download will use the RTMP protocol."""
         self.to_screen(u'RTMP download detected')
 
-    @staticmethod
-    def _decrypt_signature(s):
+    def _decrypt_signature(self, s):
         """Decrypt the key the two subkeys must have a length of 43"""
         (a,b) = s.split('.')
         if len(a) != 43 or len(b) != 43:
-            raise ExtractorError(u'Unable to decrypt signature, subkeys lengths not valid')
+            raise ExtractorError(u'Unable to decrypt signature, subkeys lengths %d.%d not supported; retrying might work' % (len(a), len(b)))
+        if self._downloader.params.get('verbose'):
+            self.to_screen('encrypted signature length %d.%d' % (len(a), len(b)))
         b = ''.join([b[:8],a[0],b[9:18],b[-4],b[19:39], b[18]])[0:40]
         a = a[-40:]
         s_dec = '.'.join((a,b))[::-1]
@@ -485,11 +485,15 @@ class YoutubeIE(InfoExtractor):
 
         try:
             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
+            if not mobj:
+                raise ValueError('Could not find vevo ID')
             info = json.loads(mobj.group(1))
             args = info['args']
-            if args.get('ptk','') == 'vevo' or 'dashmpd':
-                # Vevo videos with encrypted signatures
-                self.to_screen(u'%s: Vevo video detected.' % video_id)
+            # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
+            # this signatures are encrypted
+            m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
+            if m_s is not None:
+                self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
                 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
         except ValueError:
             pass
@@ -752,3 +756,45 @@ class YoutubeUserIE(InfoExtractor):
         urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
         url_results = [self.url_result(url, 'Youtube') for url in urls]
         return [self.playlist_result(url_results, playlist_title = username)]
+
+class YoutubeSearchIE(SearchInfoExtractor):
+    """Information Extractor for YouTube search queries."""
+    _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
+    _MAX_RESULTS = 1000
+    IE_NAME = u'youtube:search'
+    _SEARCH_KEY = 'ytsearch'
+
+    def report_download_page(self, query, pagenum):
+        """Report attempt to download search page with given number."""
+        self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
+
+    def _get_n_results(self, query, n):
+        """Get a specified number of results for a query"""
+
+        video_ids = []
+        pagenum = 0
+        limit = n
+
+        while (50 * pagenum) < limit:
+            self.report_download_page(query, pagenum+1)
+            result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
+            request = compat_urllib_request.Request(result_url)
+            try:
+                data = compat_urllib_request.urlopen(request).read().decode('utf-8')
+            except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
+                raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
+            api_response = json.loads(data)['data']
+
+            if not 'items' in api_response:
+                raise ExtractorError(u'[youtube] No video results')
+
+            new_ids = list(video['id'] for video in api_response['items'])
+            video_ids += new_ids
+
+            limit = min(n, api_response['totalItems'])
+            pagenum += 1
+
+        if len(video_ids) > n:
+            video_ids = video_ids[:n]
+        videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
+        return self.playlist_result(videos, query)